A few vagrant installations for kubernetes on VirtualBox or Parallels.

Note:

• These setups are NOT meant for a prod env.
• Just meant for practice and stuff you know :-)

These clusters can be used to practice for the certification programs.

The following cluster configurations are supported at this time:

Note that these clusters will not work next to each other as they use the same network setup and names. Choose what you want to do or practice and choose one of the setups. You are of course free to change this. I would appreciate a pull request then :-).

If you already had a cluster running and now want another one, just remove the first and start the other (see below)

All scripts have been made with a Mac in mind, but will probably work just fine on a linux based computer, and easily adjusted to work with windows. The assumption here is that if you work with k8s you will probably know something of computers 😄.

• VirtualBox installed
• Vagrant installed
• Parallels plugin installed for vagrant (vagrant plugin install vagrant-parallels)
• enough space and memory
• run mac-virtualbox-network.sh because that fixes a virtualbox network thing on the mac

Descriptions below...

It is always good to look at the scripts themselves :-)

NOTE: All commands are assumed to be performed from the root of the project folder

To create a simple k8s virtualbox cluster just clone this repo and:

• First make sure that config.vm.box = in the Vagrantfile points to a box for your chipset.

# cd <cluster_name> e.g.:
cd k8s-docker
#If you use virtualbox
vagrant up --provider=virtualbox --no-parallel
#or if you use Parallels
vagrant up --provider=parallels --no-parallel

Now be patient... the example setup (1 master, 1 worker) can easily take up to 20 minutes (or longer) to complete depending on you internet connection and the speed of your machine

To scale the amount of worker nodes you can change the following line in the Vagrantfile

1.upto(1) do |i|
  # change the upto number to the amount of workers you need/want

The k8s-docker setup, conforms to the cluster used by the LFD259 kubernetes-for-developers by the Linux foundation as of 2021 Change the upto(1) number to a higher number if you want more workers. Make sure you have enough memory on your host machine :-). Every worker will be assigned 2Gb of RAM. The Master node will be assigned 4Gb of RAM.

cd <project>
./master

cd <project>
./worker <worker_number>
#e.g.
./worker 1

./down

Note that this script takes a max of 5 worker nodes into account.

./up

Note that this script takes a max of 5 worker nodes into account.

Note that all will be destroyed!

# cd <cluster_name> e.g.:
cd k8s-docker
vagrant destroy [-f]

• Goto the Virtualbox VM folder and remove the existing image folder(s)

master: 	[WARNING IsDockerSystemdCheck]: detected "cgroupfs" as the Docker cgroup driver. The recommended driver is "systemd". Please follow the guide at https://kubernetes.io/docs/setup/cri/

• Just ignore

• ssh into your master node
• ifconfig|grep 'inet 192.168.10.1'|awk '{print $2}'|xargs echo -n to retrieve your master node ip
• kubeadm token list -o jsonpath='{.token}' to retrieve the token
• now sha256 it

openssl x509 -pubkey -in /etc/kubernetes/pki/ca.crt \
 | openssl rsa -pubin -outform der 2>/dev/null \
 | openssl dgst -sha256 -hex \
 | sed 's/ˆ.* //'

• lets combine the above commands to recreate the complete join command...

echo "sudo kubeadm join $(ifconfig|grep 'inet 192.168.10.1'|awk '{print $2}'|xargs echo -n):6443 --token $(kubeadm token list -o jsonpath='{.token}') --discovery-token-ca-cert-hash $(openssl x509 -pubkey -in /etc/kubernetes/pki/ca.crt | openssl rsa -pubin -outform der 2>/dev/null | openssl dgst -sha256 -hex | sed 's/ˆ.* //'|awk '{print $2}')"

• now copy the resulting command en perform it on your worker node(s)